ABSORPTION CHANGES IN

BACTERIAL CHROMATOPHORES

IRWIN D. KUNTZ, JR., PAUL A. LOACH, and MELVIN CALVINFrom the Department of Chemistry and Lawrence Radiation Laboratory, University ofCalifornia, Berkeley. Dr. Loach's present address is the Department of Chemistry,Northwestern University, Evanston

ABsAcr The magnitude and kinetics of photo-induced absorption changesin bacterial chromatophores (R. rubrum, R. spheroides and Chromatium) havebeen studied as a function of potential, established by added redox couples. Nophotochanges can be observed above +0.55 v or below -0.15 v. The loss ofsignal at the higher potential is centered at +0.439 v and follows a one-electronchange. The loss of signal at the lower potential is centered at -0.044 v and isalso consistent with a one-electron change. Both losses are reversible. A quantita-tive relationship exists betw%een light-minus-dark and oxidized-minus-reducedspectra in the near infrared from +0.30 to +0.55 v. Selective treatment of thechromatophores with strong oxidants irreversibly bleaches the bulk pigments butappears to leave intact those pigments responsible for the photo- and chemically-induced absorption changes. Kinetic studies of the photochanges in deaeratedsamples of R. rubrum chromatophores revealed the same rise time for bandsat 433, 792, and 865 m,u (tj = 50 msec.). However, these bands had differentdecay rates (t1 = 1.5, 0.5, 0.15 sec., respectively), indicating that they belongto different pigments. Analysis of the data indicates, as the simplest interpreta-tion, a first-order (or pseudo first-order) forward reaction and two parallelfirst-order (or pseudo first-order) decay reactions at each wavelength. Theseresults imply that all pigments whose kinetics are given are photooxidized andthe decay processes are dark reductions. These experiments are viewed assupporting and extending the concept of a bacterial photosynthetic unit, withenergy migration within it to specific sites of electron transfer.

Photo-induced absorption changes in biological material were first clearly observedby Duysens ( 1), who suggested their relationship to photosynthesis. Subsequently,more detailed studies have indicated that the photoinduced absorbance changesin bacterial systems show high quantum efficiencies (2, 3), low temperature re-versibility (4), and quenching by strong oxidants (5). These results, together withthose concerned with primary events in green plant material (6-8), have sub-stantiated the earlier concepts of photosynthesis as involving the early conversionof light energy into separated oxidants and reductants (9). Since oxidized and

227

reduced moieties are formed very early in the energy conversion process, a care-ful systematic study of the redox dependence of light-induced changes might re-sult in the identification of the participating pigments and indicate how they inter-act with one another.

If some or all of the redox couples composing the photosynthetic apparatusequilibrate with the potential established in the external medium by a suitable redoxbuffer, the effects of light might be expected to be seen as transient departuresfrom the redox equilibrium. For those systems in which the equilibrium oxidation-reduction ratio is essentially completely in favor of that oxidation state towardwhich light drives it, the light could not produce a noticeable change. An addi-tional source of information will be the rate of departure from equilibrium and therate of return.The present discussion gives the details of first results, published in part earlier

(10, 11), and of further studies concerned with the redox dependence of light-induced absorption changes from 350 to 950 mp.

In principle the kinetics of the light-produced absorption changes should pro-vide a considerable amount of information concerning the type and sequence ofreactions in the energy conversion process. In practice, the small signals, the lackof control over the pigment concentrations within a photosynthetic system, andthe complexity of the reactions have limited the amount of information readilyavailable. New instrumentation, coupled with control of the redox environment,has permitted a study of some of the kinetic interactions.

EXPERIMENTAL

Materials. Chemicals, special reagents, and the preparation of chromatophoresfrom Rhodospirillum rubrum, Rhodopseudomonas spheroides, and Chromatium used inthese studies have been described previously (12).

Methods. All redox experiments reported in detail have been conducted in thepresence of a 10- to 1000-fold molar excess of the external redox couple over the esti-mated concentration of the photoactive pigments. Values of potential are measured at22 + 20C and pH 7.4. With these large excesses of redox "buffers" the electrodes re-sponded immediately to small changes in potential (0.5 mv), and adjustments to desiredvalues were made quite easily.

For interpretation of the data obtained, it has been assumed that the relatively highconcentrations of redox buffers do not complex with the pigments being investigated tosuch an extent that the pigment properties are changed. Where possible a number ofredox couples have been used to cover the same range of potential, and experimentswere carried out to determine the effect of reagent concentration and the effects of lighton the couples.

For measurements in the absence of air, apparatus and techniques patterned after thosedesigned by Harbury (13) were used with the following modifications. A rectangular,four-sides-clear quartz cuvette was cemented (de Khotinsky's cement) to pyrex tubing,as illustrated in Fig. 1. From standard taper 2 a bridge was connected to a second cuvettewhich resided in front of the reference beam of a Cary spectrophotometer. The two

BIOPHYSICAL JOURNAL VOLUME 4 1964228

FIouRE 1 Apparatus for anaerobic redox experiments. (1) 10/30 standard taperfor sample cell (5) or microburette; (2) 10/30 standard taper for connection to ref-erence cell or for microburette; (3) four-sides-clear rectangular quartz cuvettes; (4)bridge to reference cell; (6) main titration vessel with water jacket.

cuvettes were separated, when placed in the spectrophotometer compartment, by meansof a black metal septum. With this apparatus both the reference and the sample materialcould be adjusted under identical conditions and light-minus-dark spectra observed.Chemical difference spectra could also be obtained with a minimum of manipulation ofthe sample. Also shown in Fig. 1 is a cell in standard taper 1 which could be deaeratedand filled with a sample of the material being investigated. The contents of this cell couldthen be examined by means of flashing-light spectrophotometry to be described below.

Instrumental. The light-dark difference spectra and the spectral changes at-

I. D. KUNTz, JR., P. A. LOACH, AND M. CALviN Bacterial Chromatophores 229

tendant to chemical titrations were measured on a Cary Model 14M spectrophotometer(Applied Physics Corp., Monrovia, California) equipped with a scattered transmissionattachment (Model 1462). The sample compartment was modified to admit a beam oflight which illuminated the sample at right angles to the monochromator beam. The ex-citation source was a 500 watt tungsten projection bulb. Appropriate filters preventedthe exciting beam from reaching the photomultiplier. Table I shows the filter andphotomultiplier combinations employed for different regions of the spectrum.

TABLE I

FILTERS AND PHOTOMULTIPLIERS USED IN SPECTROMETERS

Detecting Exciting Detecting Excitingwavelengths wavelengths Photomultiplier beam filters*,: beam filterst,§

Cary spectrophotometer650-1100 400-500 Dumont 6911 2403 4600, 9782380-620 650-900 RCA 6217 9782 2403260-420 750-900 Dumont 7664 9863 2600

Kinetic spectrophotometer750-1100 580-700 Dumont 6911 5031 (2) 3480

4600 (2), IFI l350-500 580-700 Dumont 6292 5031 (2) 3480

4600 (2)IFI

607 650-700 Dumont 6292 IFS IFI2030

*Placed in front of photomultiplier.tAll numbers refer to Corning color glasses.§Cary experiments included a 5 cm water filter in the exciting beam.I Baird-atomic broad band interference filter (transmits 580-720 m.").lBaird-atomic narrow band interference filter (transmits 607 ± 6 mnj).

An expanded scale slide-wire (0 to 0.2 OD unit full deflection) was as routine usedfor measuring difference spectra. Instrument controls ("slit control" and "dynodesetting") were adjusted to minimize the noise level, consistent with the resolution re-quirements of the experiments. Slit widths of up to 1 mm were found acceptable.The peak-to-peak noise under normal operating conditions was 0.0005 OD unit when

using the blue-sensitive photomultipliers. An increase in the noise (0.002 to 0.003 ODunit) and a base line shift (0.001 to 0.010 OD unit) occurred as the exciting light wasturned on for measurements above 650 mfi. These effects were not observed at low lightintensities and probably result from infrared fluorescence or scattering of the actiniclight. Since the base line shift is independent of wavelength, it was easily corrected forby selecting a wavelength at which no absorption change occurs. The precision of theseexperiments was consistent with the noise levels described.We performed the kinetic measurements on a rapid-response spectrophotometer con-

structed specially for this purpose. The instrument employed the technique of continuous

BIoPHYsIcAL JoURNAL VOLUME 4 1964230

averaging to provide both fast response and high sensitivity. This method has beendiscussed in the recent literature (14).

Fig. 2 shows a block diagram of the kinetic spectrophotometer. It operates as follows:The light from the monochromator passes through the sample and falls on a photomulti-plier. A modulated actinic beam (a neon light) illuminates the sample. Reversible changesin the absorption of light by the sample thus appear as a modulation on the transmittedintensity of the detecting beam. The photomultiplier output, then, consists of a DC signal

SPECTROMETER - BLOCK DIAGRAM

MONOCHROMATOR SAMPLE AMPUFIERS MULTISCALERCOMPARTMENr

SAMPLE CUVETTE

BAUSCHANDTETOI MNMTN

E3 PM TYP

ACTINIC TRIGGERM,/ TO X-Y RECORDERLAMP ~TRIGGER -[ jMP CIRCUIT

FIGURE 2 Block diagram for rapid-response spectrometer.

(the transmitted intensity) and a small AC signal (repetitive absorption change). Thealternating signal is amplified while the DC signal is removed by AC coupling or DC SUp-pression. The amplified signal now consists of a pulse train whose voltages are propor-tional to the periodic fluctuations in the transmitted intensity from the sample. For smallchanges in transmitted intensity the fluctuations are also proportional to the changes inthe optical density (AOD).The pulses are fed into a multiscaler unit with a built-in voltage-to-frequency converter.

The time base of the multiscaler is held in a fixed (but adjustable) phase relation to theincoming signal. After sufficient repetition to improve the signal-to-noise ratio the processis stopped and the stored signal is read out on an X-Y recorder.

Further details of the components and of the operation of the spectrometer are givenbelow:1. Monochromator: A Bausch & Lomb grating monochromator (model 33-86-45) using

a 500 watt tungsten projection bulb was operated with a 3 to 10 mnLt bandpass. Thedetecting beam is adjusted to a low enough intensity to prevent appreciable excitationof the sample.

2. Actinic lamp: The neon lamp used to induce the photoabsorption changes was kindlymade available for our use by Dr. L. Piette of Varian Associates, Palo Alto. Most ofthe details of its circuitry are in the literature (15). Its ease of modulation, rapid riseand decay times (~5 ,usec.), and convenient spectral lines make it a very usefulsource. One disadvantage is the relatively low intensity (ca. 10 quanta/sec.-cm').

I. D. KUNTZ, JR., P. A. LOACH, AND M. CALviN Bacterial Chromatophores 231

3. Photomultipliers and filters: The filters and photomultipliers used for the kineticmeasurements are given in Table I. The filters were chosen to supply the same incidentexcitation wavelengths and intensity for the detection wavelengths of greatest interest.All the photomultipliers used are of the 2 inch end-window variety. They have con-ventional base circuitry with approximately microsecond response time.

4. Amplifiers: Tektronix type D high-gain preamplifiers were employed. They werepowered by a Tektronix type 127 power supply and required a line regulator for bestperformance. Typical gains were 100 to 300, providing the 6 v (maximum) input tothe multiscaler. The preamplifiers were ordinarily Dc-coupled to avoid distortion ofthe rather slow waveforms. A simple battery circuit was used to bias out any DCvoltages at the input to the preamplifiers.

5. Multiscaler unit: A commercial unit, the Computer of Average Transients (MnemotronCorp., Pearl River, New York), model 400, satisfactorily met our requirements ofmillisecond response time and 1 per cent accuracy. Internal noise, although notnegligible, was usually well below the noise at the input. At times, coherent internalnoise proved troublesome. We employed an external triggering unit to remove thisdifficulty. The final trigger circuit used is shown in block form in Fig. 3.

TRIGGER CIRCUIT

|TYPE 162 'TYPE 161 TYP 162

SAWTOOTH " PULSER MULTISCALER ,PULSER

GENERATOR,

EXTERNAL TRIGGERUNIT TO ACTINIC

LAMPFiGuRE 3 Trigger circuits used in rapid-response spectrometer. The external triggerunit produces a -1.5 v, 0.1 msec. square pulse to start the multiscaler sweep. The multi-scaler then triggers a second Tektronix type 162 unit which pulses on the actiniclamp. The lamp remains on for the duration of the pulse (100 ,ssec. to 10 sec.).

Performance. This instrument has an over-all rise time faster than 0.1 msec. Themajor noise source (the shot noise in the detecting beam) produces about 10-' OD unitsof noise for typical operating conditions, that is, 2 to 10 minute experiments. Repetitionrates were determined by the decay time of the biological material. Rates of 4 cps to0.03 cps have been used. An interesting feature of the system is that a moderate amountof fluorescence can be conveniently compensated for by turning off the detecting beam(thus removing all of the absorption changes but not the fluorescence) and running themultiscaler for an equivalent length of time in the subtract mode.The over-all performance of the spectrometer is limited by the reproducibility of the

biological responses. Magnitudes of the absorption changes were reproducible to a fewper cent. The rates and rate constants given are probably accurate to 20 to 30 per cent.

BIOPHYSICAL JoURNAL VOLUME 4 1964232

RESULTS

Magnitude of Reversible Photoabsorption Changes as a Function of RedoxPotential.

1. High Potential. Samples of the deaerated chromatophores were treatedwith mixtures of K3Fe(CN)6 and K4Fe(CN)6 to cover the potential range of +0.30to +0.55 v." Representative absorption spectra and light-dark difference spectrafor control samples (EA = +0.35 v)2 are shown in Figs. 4 to 6. These results arein relatively good agreement with those of other workers (1, 3, 17).The oxidized-minus-reduced and the light-minus-dark signals at 792, 810, and

865 mpu were measured at each potential. As shown in Fig. 7 and Table II, raising

TABLE II

QUANTITATIVE COMPARISON OF LIGHT-DARK ANDCHEMICALLY-INDUCED SPECTRAL CHANGES*

AOD72 + AODs1i AODse

Chemically Total active Chemically Total activeEA, Photooxidizable oxidized pigment Photooxidizable oxidized pigment

0.377 0.051 0.005t 0.056 0.041 0.004t 0.0450.39. 0.043 0.014 0.057 0.032 0.009 0.0410.41k 0.036 0.019 0.055 0.032 0.016 0.0480.432 0.032 0.026 0.058 0.026 0.022 0.0480.469 0.016 0.039 0.055 0.015 0.036 0.0510.49k 0.008 0.045 0.053 0.008 0.036 0.044

*Data not corrected for irreversible selective or general bleaching.I:Arbitrary zero assigned on basis of 90 per cent light changes at 0.38 v.

the potential through the region of +0.4 v removes the light-dark signals, replac-ing them with the same bands in the oxidized-reduced spectra. The samples werethen reduced in a stepwise fashion with sodium dithionite to demonstrate the re-versibility of the effects. The midpoint of both the disappearance of the photo-changes and the chemical titrations for each wavelength studied was +0.439 v.Fig. 7 also shows the theoretical curves for one- and two-electron transitions withEm = +0.439 v. The data are consistent with a one-electron transition occurringin a key pigment(s) with a midpoint of +0.439 v. Within the limits of the experi-ment there is quantitative agreement between the amount of material available for

1 In the absence of externally added redox couples, the oxidation-reduction potential of eachsample could be measured and the system seemed to be somewhat buffered at the measuredvalue, but the couple(s) equilibrating with the electrode is unknown.2 The symbols E, and E. are used as suggested by Clark (16).

I. D. KuNTz, JR., P. A. LOACH, AND M. CALViN Bacterial Chromatophores 233

reversible photochemistry and that available for reversible oxidation (Table II,

columns 4 and 7).In the above experiment potassium iridic chloride could replace potassium fer-

ricyanide as oxidant, and potassium ferrocyanide could replace sodium dithioniteas reducing agent. Results to be described below show that changes in the kineticsare not responsible for the loss of signal.

0.7A

R. rubrum 8800.606 CHROMATOPHORES

0.5

0.4

0.030

0.020

0.010

a0

.1

-0.01

-0.02

-0.03

400 500 600 700 800 900WAVELENGTH (nw4

0.060

0.040

0.020

I 0

0

0.020

0.040

0.060

FIGURE 4A Absorption spectrum of R. rubrum chromatophores suspended in 0.01M phosphate buffer, pH 7.13; 1 cm cuvette.FIGURE 4B Light-induced absorption changes in R. rubrum chromatophores whoseabsorbance was 2.2 at 880. Note that different excitation wavelengths were used aboveand below 650 m,u (Table I). The absorption scale below 650 m,u is expanded two-fold. Air removed; 1 cm cuvettes.

BIoPHYscAAL JouRNAL VOLUME 4 1964

w .** * *Il* * -

., I s.

B

433 116

0 1-

385 605 1 810

0 I.~~~~~~~~~~~~~~~

I ,. .

, I . , , . I

Il

234

710 765 792

U IA

1.40 R spnero,aesCHROMATOPHORES

1.20375

1.00

476

0.80 510o ~~~4

0O060

0.40590

0.20

400 500 600 700 800 900

0.0.B 522

QLO6 _1Q065 IOD5 I

0.04 4 I

0.03

iO.OSl 428p460p8

-OD1~412

44460-0.02. 470

-0.03 I8

510I-0.04I

WAVELENGTH (mp)

FIGURE SA Absorption spectrum of R. spheroides chromatophores suspended in0.05 m phosphate buffer, pH 7.4; 1 cm cuvette.FIouRE SB Light-induced absorption changes in R. spheroides chromatophoreswhose absorption is given above. Air removed; 1 cm cuvettes.

849

4

. I . a I I , I I I I I I 0 I - . . v I . . . . I . 9 . . .

A

I I I I

O 0.500

0.40 887

Q30590

0.20

0.10

400 500 60 700 800 900

788

0.01 450

485 590 70 /

0

556-0.01I

-0.0288

400 500 600 700 800 900WAVELENGTH (m)

FIGURE 6A Absorption spectrum of Chromatium chromatophores suspended in0.05 M phosphate buffer, pH 7.4; 1 cm cuvette.FIGURE 6B Light-induced absorption changes in Chromatium chromatophores whoseabsorption is given above. Air removed; 1 cm cuvettes.

BIOPHYSICAL JOURNAL VOLUME 4 1964236

2. Intermediate Potential. Experiments conducted through the range ofpotential from +0.05 to +0.35 v showed no variation in the spectra of absorbancechanges.

3. Low Potential. The reversible light-induced absorbance changes in R.rubrum chromatophores are not observed if the redox level of the solution is below-0.15 v. Fig. 8 shows the effect of lowering the potential through the region of+0.02 v to -0.1 v. The oxidized forms of indigo tetra- and trisulfonic acid werepresent at 2 X 10-sM as "buffers" in the potential ranges +0.02 to -0.06 and -0.06to -0.10 v, respectively. The potential was lowered slowly by the addition of re-ducing agent, reduced indigodisulfonic acid, prepared separately by 95 per cent re-

,~ ~ ~ ~ ~~ ~-'.4f(j .*&,..,,i ',..§t¢

FI*uRE 7 Variation with "high" potential of the light-induced absorption changes at865, 810, and 792 m/j& in chromatophores from R. rubrum. 0.05 m phosphate buffer,pH 7.4. Absorption at 880 mju was 2.0. Total concentration of ferricyanide and ferro-cyanide varied from 5 X l01 m to 2 X 10-' m. Reducing agent, 0.01 m Na&S204.After removal of oxygen, enough KtFe(CN). was added to make it 5 X 10 Mt0, A, 0, experimental points taken at 865, 810, and 792 mj after increaKaFe(CN),o concentration to increase potential. 0, A, U, experimental points takenat 865, 810, and 792 mg upon addition of increasing amounts of Na,S2,0 to lowerthe potential, beginning after all KX,e(CN)a had been added. The solid and dashedlines were obtained by use of the equation E,. = E. + RT/nJ ln(m.s. - l.s.)/ls.with n = 1 and 2, respectively, and E. = +0.439 v, m.s. = mai light-inducedsignal, observed between +0.35 to +0.30 v, l.s. = light-induced signal observed atpotential reported.

L. D. KUNTZ, JR., P. A. LOACH, AND M. CALVI Bacterial Chromatophores 237

duction with sodium dithionite. The absorbance changes at 792 and 865 n* dis-appeared in the same relative proportions as the potential was lowered. Unlike theferricyanide treatment, this reduction does not replace the light-induced signalswith chemically-produced counterparts. The only pronounced change in the visibleor near infrared spectrum is the reduction of an iron prophyrin complex (probablythe RHP (Rhodospirillum heme protein) of Kamen and Bartsch, 18). The data ofFig. 8 are compared with theoretical curves representing one- and two-electrontransitions. The data are consistent with a one-electron change with a midpoint of

,' 090 R. rubrum / o /

CHROMATOPHORES / a80 _

70 _

60 /8 .

50x

"40 _ 0

30 _

200 I

10_/oco ,

-0.14 -010 -Q06 -002 0 0.02 0.06Eh (VOLT)

FoLuE 8 Variation with "low" potential in the light-induced absorption changesat 810 and 792 m,u in R. rubrum chromatophores. 0.05 M phosphate buffer, pH 7.4.Absorption at 880 m,u was 1.22. Total concentration of reductant used varied from5 X 1OV M to 6 x 10- M. Reducing agent, reduced indigodisulfonic acid. The dyesindigotetrasulfonic acid and indigotrisulfonic acid were present at approximately2 X 10- M concentrations to act as redox buffers. 0, experimental points obtainedfrom the sum of A OD at 792 and 810 mn as recorded by the Cary spectrophotometerusing very high light intensity; points were recorded sequentially as the potential waslowered. 0, experimental point obtained as circles except KFe(CN). was added to-0.12 v sample. i\ experimental points taken from a separate experiment in whichindigotetrasulfonic acid (1 X 101 M) was the only dye present. 5], experimentalpoints obtained from steady-state values at 792 m,u using the kinetic spectrometer;these data have been corrected for changes in decay rate over this potential range(Fig. 16) as noted in the text. The solid and dashed lines were obtained by use of theequation Ea = E. + RT/nJ In l.s./(m.s.- I.s.) with Em = -0.044 v and n = 1 and2, respectively. As in Fig. 7, m.s. = maximum signal observed between + 0.35 and+0.30 v and l.s. = the light-induced signal observed at the potential reported. Anaero-bic conditions.

BIOPHYrSICAL JOURNAL VOLUME 4 1964238

0.Ic0.1 I I I I . . A I - . I I ,..II .I I IA K2IrCI6-TREATED

R. rlbrum0.08 CHROMATOPHORES

0.06

Co \ 540 595 640 715 8080.04

0.02

I .. I I 1 ..t I 1 . .

400 500 600 700 800 900

0.008B

7900.004

710

0 I0

I 810-0.008

-0.012 868

-0.016

400 500 600 700 80D 900

WAVELENGTH (mj4

FIGURE 9A Absorption spectrum of KJrClv-treated R. rubrum chromatophores sus-pended in 0.05 M phosphate buffer, pH 7.4; 1 cm cuvette.FIGuRE 9B Light-induced absorption changes in the sample whose absorbance isgiven above; 1 cm cuvettes.

L. D. KuNTz, JR., P. A. LOACH, AND M. CALVIN Bacterial Chromatophores 239

o 0.080~~~~~~~~~~~~~0

0.06 685

0.04 5 765

0.02

400 500 600 700 800 900

0.012

Q008 _ 792

0.004 4300~~~~~~~~~~~20

a0.004 /5

-0.008 _585 :V

-0.012t 81

-0.016186

-0.020_I

400 500 600 700 800 900WAVELENGTH (mIA)

FIGuRE 10A Absorption spectrum of K2IrCl0-treated R. spheroides chromatophoressuspended in 0.05 M phosphate buffer, pH 7.4. Excess K,Fe(CN). present (approxi-mately 0.05 M); 1 cm cuvette.FIGuRE lOB Light-induced absorption changes in the sample whose absorbance isgiven above; 1 cm cuvettes.

BIOPHYSICAL JOURNAL VOLUME 4 1964240

o 0.08760

0.06 600W

0.04

0.02

400 500 600 700 800 9000 . . . I I I

B I 797

+0.01 I

411 A

00 A0 ~~~525 56

..0.0l 42

810 83

400 s.500 600 700 800 900WAVELENGTH (mj)

FiouRE 11A Absorption spectrum of KjIrCl-treated Chromatium chromatophoressuspended in 0.05 M phosphate buffer, pH 7.4. Excess K,Fe(CN). present (approxi-mately 0.05 M); 1 cm cuvette.FIGuRE 1 lB Light-induced absorption changes in the sample whose absorbance isgiven above; 1 cm cuvettes.

I. D. KUNTZ, JR., P. A. LOACH, AND M. CALvi Bacterial Chromatophores 241

-0.044 v. Again, results to be described below show that changes in the kinetics arenot alone responsible for the loss of signal.The effects of 10-5 M concentrations of other strong reducing agents (sodium

dithionite, semiiquinone of methylviologen, and reduced phenosafranine) were eachexamined on separate samples of well deaerated R. rubrum chromatophores. Ineach case, no light-induced signals could be observed in the presence of these re-ducing agents.

Separation of Photoactive Pigments from Bulk Pigments by Use of StrongOxidants

Irreversible oxidation of large amounts of the long wavelength absorbing pigmentsin the presence of an excess of strong oxidant (12) leaves a material which stillhas absorption changes of the same magnitude as the original samples. Figs. 9 to11 compare the light-dark changes and absorption spectra of K2IrC-treatedchromatophores prepared from R. rubrum, R. spheroides, and Chromatium. Thesimilarity of the absorbance of these materials in the near infrared is as strildng

0.75

; ; o/ - 9~~0 mFL* 43Smp&

FIouRE 12 Variation with wavelength in the rise kinetics of the light-induced ab-sorption changes in R. rubrum chromatophores at 433, 792, 810, and 865 m&. Forpurposes of comparison signal heights are normalized and all signals are shown aspositive. The actinic lamp was on for 0.4 sec. and off for 3.6 sec. Anaerobic condi-tions; 10-' M KFe(CN). present to keep Eh, = +0.35 v; 1 cm cuvette; 0.01 M phos-phate buffer, pH 7.4; OD at 880 mA, 0.70.

BIOPHYSICAL JOURNAL VOLUME 4 1964242

as the dissimilarity among the absorbance of the original chromatophores. Notealso the loss of the light-dark carotenoid change in R. spheroides (compare changesat 444, 460, 470, 487, 510, and 522 mp in Figs. SB and lOB), and the participa-tion of what is probably an iron prophyrin complex in the light-induced changes ofChromatium (compare similarities at wavelengths 410, 423, and 556 m,z in Figs.6B and 1iB). Higher incident light intensity is required to produce the same sizeabsorption change in these K2IrCI-treated systems as in the unmodified material; thisseems to reflect the loss in absorption rather than a major change in quantum ef-ficiency.

If the K2IrCI8-treated material from R. rubrum is chemically titrated in the darkwith ferricyanide, absorbance changes identical to those obtained by the action oflight can be observed. Thus, it appears that the bulk pigments are not essentialcomponents in either the light-produced absorption changes or the chemically-in-duced changes.

Kinetics

Rise and decay times were studied as a function of wavelength, light intensity, andpotential.

1.0

a't R. rubrumCHROMATOPHORES

0.75

z

0.50 43p

a 792 ml.

0 1.0 2.0 3.0 4.0 5.0

L t TIME (SEC.)LIGHTON OFF

FoGuRE 13 Variation with wavelength in the decay kinetics of the light-induced ab-sorption changes at 433, 792, 810, and 865 m,in in R. rubrum chromatophores. Con-ditions as in Fig. 12. The bars represent the total estimated uncertainties including baseline drift and sample variability. The actinic lamp was on 0.5 sec. and off 7.5 sec.

I. D. KuNTz, JR., P. A. LOACH, AND M. CALVIN Bacterial Chromatophores 243

Variation of Kinetics with Wavelength. Typical rise and decay curves at433, 792, 810, and 865 m/ for deaerated samples (Eh - +0.35 v) are shown inFigs. 12 and 13. The rise times are inversely related to the absorbed intensity fromthe actinic source. Hence, different sources, high optical densities, or the presenceof colored reagents can change the rise constants. The decay constants are notparticularly sensitive to these effects. Although the time for half-rise is nearly thesame for each band measured, significant differences occur in the decay kinetics.The rise and decay kinetics are complex. The simplest interpretation of the datais a first-order (or pseudo first-order) forward reaction and two parallel first-order(or pseudo first-order) decay (or back) reactions (Figs. 14 and 15). Rate con-stants for the first-order reactions are compared in Table III.

These different decay kinetics at different wavelengths clearly indicate that theabsorption changes studied belong to at least three different pigments (Fig. 13).

i.595% OF

STEADY- STATE,./VALUE

1.0

0

R. rubrum0.5 // CHROMATOPHORES

RISE TIME,Kf433 mtL

0 100 200TIME (MSEC.)

FiGuRE 14 Analysis of rise kinetics for 433 mnu as shown in Fig. 12. XO = maximumobserved steady-state absorption change, X = absorption change at any given time.The solid line is the theoretical plot for a first-order reaction with Kr = 16 see'.Other wavelengths show similar behavior.

Variation of Kinetics with Light Intensity. A further study of the kineticorder of the decay reactions involves changing the steady-state level of photo-products by variation of the incident light intensity. A fourfold change in the steady-state level produced less than a 20 per cent change in the time for half-decay. This

BIOPHYSICAL JOURNAL VOLUME 4 1964244

w VI._z R. rubrum< Ma CHROMATOPHORE S

50 792 mbo

0~~~~~0

a.

FIUR 1 nayssofdea10etc deof .thih-nucda4rtoncagsa

0

792 mJu in R. rubrum chromatophores shown in Fig. 13. The vertical scale is loga-rithmic. Two decay processes are shown by solid lines (K,1 and K,.,). Twhe computeddecay pattern (dotted line) was obtained from

XaGll = ao exp (-wK1t)+ p exp ( KB2t)where X = concentration of photoproduced species, K,1 and K,., are first-order decayconstants 2.0 and 0.46 sec.-1, respectively. ,s is the extrapolated Y intercept for thleslow decay process. a = total initial signal minus p. t is time measured from whenthe lamp went off.

observation strongly supports the assignment of one or more first-order or pseudofirst-order decay reactions to each of the three different pigments.

Variation of Kinetics with Potential. While the rate constant for the for-ward reaction, K,, is nearly independent of potential in the regions studied, thedecay constant, KB1, is a function of redox potential. Fig. 16 shows the variationof the first-order decay constant K31 for the changes in the 865 m,u band in the

TABLE III

RATE CONSTANTS FOR PHOTOABSORPTION CHIANGES*(assuming a first-order forward reaction (K,.) and two parallel

first-order back reactions (KB1, KB2))

Wavelength Kf (sec.-) KB1 (sec.-') KB2 (sec.C1)

433 16.3 4 0.3 0.9 -t 0.2 0.2 0.06792 15.940.3 1.740.3 0.4 t0.1810 16.1 40.3 1.940.3 0.440.1865 16.3 0.3 4.8 4 0.6 0.9 + 0.2

*Deaerated samples, Eh " +0.35 v.

I. D. KUNz, JR., P. A. LOACH, AND M. CALvIN Bacterial Chromatophores

I t_k

245

A t

R. rubrum CHROMATOPHORES0 DECAY CONSTANT, 865 miL

15

:b

0@ 10

00

5 0

00 0 o O o

0 I I I I I I I I I I-0.10 0 0.10 0.20 0.30 0.40

Eh (VOLT);

FiouRE 16 Variation with potential of the decay rate constant (Km,) of light-in-duced absorption changes in R. rubrum chromatophores. Experimental conditionsas given in Figs. 7 and 8.

potential range of -0.15 to +0.5 v. The rate constant for the 792 m,t band showedsimilar behavior. Generally, the faster of the two decay constants increases withdecreasing potential. The relative amount of material decaying by the fast routealso increases as the potential is lowered. The difficulty in measuring the constantfor the slower decay process KB2 precludes a generalization of its behavior. Therapid change in KB1 in the vicinity of -0.05 v suggests the reduction of a pigmentwhich can then interact with the photooxidized materials.The decay rates never become rapid enough to explain the loss in signals at low

potential (or, for that matter, at high potential). For the reactions described herethe steady-state concentration of the photooxidized species is given by

__ __ K ( K,./KBK OOr P+ B)Po -P+ KB l + Kf/K Po

where Po is the initial concentration of the oxidizable species, P+ is the concentra-tion of the oxidized material, Kf is the first-order forward rate constant, KB is thefast back reaction first-order rate constant. Kf = 16 sec.-", KB varies from 2 to14 sec.-' (at 792 mn). Thus, the steady-state value of P+ would fall from 89 percent of Po to 53 per cent of Po because of the increase in decay constant from 2to 14 sec-'. Experimentally the steady-state level of P+ fell to lower than 2 percent for KB = 14 sec.-1 (at El = -0.12 v). These findings support the assign-

BIOPnYSICAL JOURNAL VOLUME 4 1964246

ment of the losses in the steady-state signal to the titration of pigments required forthe appearance of the light signal rather than those hastening its disappearance.

It is of interest to note that the decay rates tend to approach the same high value(KB1 14 sec.-l) for each of the three wavelengths studied at high concentrationsof K4Fe(CN)8 (> 0.1 M).

Variation of Kinetics with Oxygen Pressure. The effect of oxygen on therise and decay kinetics was examined at two fixed potentials, +0.44 v and +0.30v.No significant effect was observed on any of the bands studied.

DISCUSSION

The experiments presented here are interpreted as characterizing a photoactiveunit composed of at least two essential pigments. Photosynthetic activity (asmeasured by the typical light-induced absorption changes) in chromatophoresfrom R. rubrum, R. spheroides, and Chromatium is dependent upon the reducedform of a pigment, P0.44 and the oxidized form of a pigment P.0.04 which appearto be capable of undergoing a one-electron oxidation or reduction, respectively.'Thus, P0.44 must be the primary electron donor, or kinetically indistinguishablefrom it, and P.0.04 may be the primary electron acceptor, or kinetically indis-tinguishable from it in these bacterial systems.4 Other interpretations of the lossof absorption changes below 0 v, such as further reduction of P0.44 or complexformation with the added redox couples, cannot be ruled out at the present time.A fundamental question is that of the relationship of PO.44 to the oxidized-minus-

reduced and light-minus-dark difference spectra. The available evidence cannotsupport a rigorous assignment of absorption bands to Po.44 but some deductionscan be made. The agreement between oxidized-minus-reduced and light-minus-dark spectra in the near infrared demonstrates that the photoabsorption changes at792, 810, and 865 mp are associated with electron donors. Thus, one or more ofthese bands could be due to Po.44. The difference in decay kinetics for the photo-absorption changes at 433, 792, 810, and 865 mp (Fig. 13) clearly indicates thatthere are at least three different molecules affected by the light. It follows that Po.44is most likely to absorb strongly at only one of these wavelengths (if at all).A tentative working hypothesis assigns to PO.44 an 865 mu band in the reduced

form. This probably is only one of several reduced bands, and no bands have beenidentified with the oxidized form since no photoinduced increases in absorbance be-tween 350 and 950 mju show kinetics to correspond with the 865 decrease.The emerging picture of a discrete active unit is nicely supported by the iridic

This notation is used to stress the midpoint potential (+0.44 or -0.04 v) for the redoxchanges and to stress our lack of knowledge about the complex.'A midpoint of +0.43 v for the light-induced absorption loss at 700 mu in chloroplasts hasbeen reported by Kok (19). The similarity of the potentials in the two systems is probably notaccidental.

I. D. KUNTZ, JR., P. A. LOACH, AND M. CALviN Bacterial Chromatophores 247

chloride experiments and Clayton's observations with pheophytinized R. spheroideschromatophores (17). The removal of the bulk pigment absorption produces onlysmall changes in the IR spectra of the photosignals and does not change their magni-tude. It thus appears that the 880 mu pigment in R. rubrum, the 850 mub and partof the 800 m/A pigment in R. spheroides, and the 880 and 800 mjb pigments inChromatium, which can be irreversibly removed by chemical oxidation, are not es-sential components in either the light-produced absorption changes or the light-produced ESR signals (12), both of which may be produced qualitatively andquantitatively in the same fashion in their absence as in their presence. Thus, thebulk pigments can only act as energy-gathering and energy-transfering systemswhich may supply the photoactive unit. This clean-cut separation of the active andbulk pigments offers considerable support to the concept of a "photosynthetic unit"(20, 1, 17, 3).

The authors wish to thank Mr. Bill Hart for construction of all glass apparatus used foranaerobic measurements and Mr. Roland Thompson for his expert electronic advice and as-sistance in the construction of the kinetic spectrophotometer. We are indebted to Dr. KennethSauer for his many helpful discussions, advice, and encouragement.

Irwin D. Kuntz, Jr., was a National Science Foundation Fellow, 1961-1963. Paul A. Loachwas a Fellow of the National Academy of Sciences-National Research Council, supported bythe Air Force Office of Scientific Research of the Office of Aerospace Research, 1961-1963.The work described in this paper was sponsored, in part, by the United States Atomic EnergyCommission.Received for publication, July 16, 1963.

REFERENCES

1. DUYSENS, L. N. M., 1952, Doctoral Thesis, Utrecht.2. OLSON, J. M., and CHANCE, B., Arch. Biochem. and Biophysics, 1960, 88, 40.3. CLAYTON, R. K., Photochem. and Photobiol., 1962, 1, 305.4. ARNOLD, W., and CLAYTON, R. K., Proc. Nat. Acad. Sc., 1960, 46, 769.5. GOEDHEER, J. C., Brookhaven Symp. Biol., 1958, 11, 325.6. WrrT, H. T., MULLER, A., and RUMBERG, B., Nature, 1963, 197, 987.7. KOK, B., and HOCH, G., in Light and Life, (W. D. McELRoY and B. GLASS, editors),

Baltimore, The Johns Hopkins Press, 1961, 397.8. SoGo, P. B., PON, N. G., and CALVIN, M., Proc. Nat. Acad. Sc., 1957, 43, 387.9. VAN NIEL, C. B., Arch. Mikrobiol., 1931, 3, 1.10. CALVIN, M., and ANDROES, G. M., in La Photosynthese, Paris, Editions du Centre National

de la Recherche Scientifique, 1963, 21.11. CALVIN, M., and ANDROES, G. M., Science, 1962, 138, 867.12. LOACH, P. A., ANDROES, G. M., MAKSIM, A. F., and CALVIN, M., Photochem. and Photobiol.,

1963, 2, 443.13. HARBURY, H. A., J. Biol. Chem., 1957, 225, 1009.14. KLEIN, M. P., and BARTON, G. W., Rev. Scient. Instr., 1963, 34, 754.15. BELL, W. E., BLOOM, A. L., and LYNCH, J., Rev. Scient. Instr., 1961, 32, 688.

248 BIOPHYSICAL JOURNAL VOLUME 4 1964

16. CLARK, W. M., Oxidation-Reduction Potentials of Organic Systems, Baltimore, Williams& Wilkins Co., 1960.

17. CLAYTON, R. K., Photochem. and Photobiol., 1962, 1, 201.18. KAMEN, M. D., and BARTSCH, R. G., Internat., Union Biochem. Symp. Haematin Enzymes,

Canberra, 1959, 419.19. KOK, B., Biochim. et Biophysica Acta, 1961, 48, 527.20. EMERSON, R., and ARNOLD, W., J. Gen. Physiol., 1932, 16, 191.

I. D. KUNTZ, JR., P. A. LOACH, AND M. CALVIN Bacterial Chromatophores 249